Bruno Wolff III writes:
> P.S. Using spgist with version 10 for the exclude index is much faster
> than using gist in 9.6. I have run the index creation for as long as
> 6 hours and it hasn't completed with 9.6. It took less than 10 minutes
> to create it in 10. For this
On Wed, Apr 05, 2017 at 12:11:09 -0600,
Rob Sargent wrote:
On 04/05/2017 12:04 PM, Bruno Wolff III wrote:
On Wed, Apr 05, 2017 at 00:05:31 -0400,
Tom Lane wrote:
Bruno Wolff III writes:
... I create both a normal gist index and
On 04/05/2017 12:04 PM, Bruno Wolff III wrote:
On Wed, Apr 05, 2017 at 00:05:31 -0400,
Tom Lane wrote:
Bruno Wolff III writes:
... I create both a normal gist index and an exclude index using the
following:
CREATE INDEX contains ON iplocation USING gist
On Wed, Apr 05, 2017 at 00:05:31 -0400,
Tom Lane wrote:
Bruno Wolff III writes:
... I create both a normal gist index and an exclude index using the
following:
CREATE INDEX contains ON iplocation USING gist (network inet_ops);
ALTER TABLE iplocation
ADD
Bruno Wolff III writes:
> ... I create both a normal gist index and an exclude index using the
> following:
> CREATE INDEX contains ON iplocation USING gist (network inet_ops);
> ALTER TABLE iplocation
> ADD CONSTRAINT overlap EXCLUDE USING gist (network inet_ops WITH &&);
>
I am trying to load a database with about 3.5 million records relating
netblocks to locations. I currently don't know whether or not any of the
netblocks overlap. If they don't, then I can simplify queries that
find the locations of IP addresses.
I create the table as follows:
DROP TABLE IF